基于RBF神经网络的感应电机自抗扰控制仿真

Simulation of Active Disturbance Rejection Control of Induction Motor Based on RBF Neural Network

感应电机在工作过程中,容易受到外界干扰源的干扰,若干扰程度严重会造成电机故障,影响电机运行状态,为了提升感应电机的自抗扰能力,提出基于RBF神经网络的感应电机自抗扰控制方法。建立感应电机数学模型,依据电机损耗分析结果,确定电机内部非线性关系;结合RBF神经网络构建感应电机的控制系统,并将该系统划分成RBF网络辨识器以及自抗扰控制器两个部分,通过二者之间的有效结合,完成感应电机的自抗扰控制。实验结果表明,使用上述方法开展电机自抗扰控制时,响应速度较快,电机电流输出与实际结果之间差距较小,说明其控制性能较好。

In order to improve the active disturbance rejection ability of the induction motor,a method of controlling active disturbance rejection of the induction motor was proposed based on RBF neural network.Firstly,a mathematical model of the induction motor was built.Based on the analysis of motor loss,the internal nonlinear relationship of the motor was determined.Then,the control system of the induction motor was constructed by RBF neural network.Meanwhile,the system was divided into two parts:RBF network identifier and active disturbance rejection controller.Finally,based on the effective combination of the two,the control of active disturbance rejection of the induction motor was completed.Experimental results show that the method has ideal response speed.The difference between the current output and the actual result is small,indicating that the control performance is better.